- Email: [email protected]
Effect of common ions on nitrate removal by zero-valent iron from alkaline soil
Journal of Hazardous Materials 231–232 (2012) 114–119
Contents lists available at SciVerse ScienceDirect
Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat
Effect of common ions on nitrate removal by zero-valent iron from alkaline soil Cilai Tang, Zengqiang Zhang ∗ , Xining Sun College of Resources and Environment, Northwest A&F University, Yangling, Shannxi 712100, China
h i g h l i g h t s The selected cations and anions enhance nitrate reduction by Fe0 in alkaline soil. Ammonium is major final product from nitrate reduction. The results prove the feasibility of using Fe0 for groundwater nitrate remediation.
a r t i c l e
i n f o
Article history: Received 29 March 2012 Received in revised form 21 June 2012 Accepted 21 June 2012 Available online 29 June 2012 Keywords: Zero-valent iron Nitrate reduction Soil and groundwater remediation Loess soil Cations Anions
a b s t r a c t Zero-valent iron (Fe0 )-based permeable reactive barrier (PRB) technology has been proved to be effective for soil and groundwater nitrate remediation under acidic or near neutral conditions. But few studies have been reported about it and the effects of coexistent ions under alkaline conditions. In this study, nitrate reduction by Fe0 was evaluated via batch tests in the presence of alkaline soil and common cation (Fe2+ , Fe3+ and Cu2+ ) and anion (citrate, oxalate, acetate, SO4 2− , PO4 3− , Cl− and HCO3 − ). The results showed that cation significantly enhanced nitrate reduction with an order of Fe3+ > Fe2+ > Cu2+ due to providing Fe2+ directly or indirectly. Most anions enhanced nitrate reduction, but PO4 3− behaved inhibition. The promotion decreased in the order of citrate > acetate > SO4 2− > Cl− ≈ HCO3 − ≈ oxalate PO4 3− . Ammonium was the major final product from nitrate reduction by Fe0 , while a little nitrite accumulated in the beginning of reaction. The nitrogen recovery in liquid and gas phase was only 56–78% after reaction due to ammonium adsorption onto soil. The solution pH and electric conductivity (EC) varied depending on the specific ion added. The results implied that PRB based Fe0 is a potential approach for in situ remediation of soil and groundwater nitrate contamination in the alkaline conditions. © 2012 Elsevier B.V. All rights reserved.
1. Introduction The loess plateau, an arid or semi-arid region, is the predominant terrain in the northwest of China, comprising an area of 6.24 × 105 km2 . Loess soil is an alkaline soil. In this area, water resource is scarce and groundwater is the predominant water source for people living, agricultural and industrial use. But as the industrial and agricultural development, nitrogen-containing industrial waste effluent discharge, excessive use of nitrogen fertilizer to improve food output for increasing population induced increasingly nitrate contamination in the groundwater [1]. An investigation on sub-aqueous nitrate contamination in Guanzhong basin (a typical loess plateau), Shaanxi Province in China, indicated that nitrate concentration in most groundwater markedly exceeded the drinking water standard (GB 5749-2006) of China (10 mg N L−1 ) [2]. Excessive nitrate in drinking water can cause cancer and other diseases [3]. Loess soil is an alkaline soil that has an average pH
∗ Corresponding author. E-mail address: [email protected] (Z. Zhang). 0304-3894/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jhazmat.2012.06.042
value above 8.0, and HCO3 − , SO4 2− , Cl− , Na+ , Ca2+ , and K+ are the major ions in the soil and groundwater [4]. Many studies have shown that nitrate reduction by Fe0 was an acid-driven and surface-mediated process, which was spontaneous under acidic condition, or near neutral condition with the help of catalysts [5–8]. Permeable reactive barrier (PRB) based Fe0 has been successfully used for in situ groundwater contamination remediation, including nitrate, under acidic or near neutral conditions [9]. But no study has been reported whether or not it could work under alkaline condition, such as in alkaline loess plateau. Granular iron reactivity is predominantly controlled by the groundwater geochemistry (e.g., coexistent ion and pH) [10,11]. The coexistent ions behaved different impacts on Fe0 reactivity depending on the target pollutants [12,13]. Fe2+ , Fe3+ and corresponding hydroxides are the oxidative products from Fe0 corrosion. Lots of previous studies showed that Fe2+ or Fe3+ enhanced Fe0 degradation of pentachlorophenol [14], chromate [15] and nitrate [16,17]. Cu(II) was usually added as a catalyst for enhancing pollutants removal by Fe0 [6,14]. Cations (Fe2+ , Fe3+ and Cu2+ ) without soil have been proved to promote nitrate reduction by Fe0 but there were no distinct effect by Ca2+ , Na+ and K+ based on the former studies [17]. Therefore, we
C. Tang et al. / Journal of Hazardous Materials 231–232 (2012) 114–119
115
copper were not detectable. All processes were performed as the methods for the examination of water and wastewater of China (4th edition). 2.2. Experiments
Fig. 1. SEM photograph of iron grains.
also investigated the effects of Fe2+ , Fe3+ and Cu2+ on nitrate reduction by Fe0 in the presence of loess soil. Moreover, the common inorganic ions such as Cl− , HCO3 − , SO4 2− , PO4 3− and potential coexistent organic ions such as citrate, oxalate and acetate have been selected to evaluate their effects on nitrate removal by Fe0 . The objective of this study was to evaluate the effects of common ions in soil and groundwater on nitrate reduction by Fe0 in the presence of alkaline soil, aiming to give some possible guidance to practical application of PRB technology for in situ removal of nitrate in the loess plateau area in the future. 2. Materials and methods 2.1. Materials Unless otherwise indicated, all chemicals used were reagent grade and all aqueous solutions were prepared with deionized water. Nitrate solution was prepared using NaNO3 . Cation and anion stock solutions were prepared with corresponding chloride and sodium salts, respectively. The iron particle (Tianjin Zonghengxing Industrial &Trading Chemical Reagent Co., China) size was approximately 50–100 m in diameter (Fig. 1), irregular in shape, with a specific surface area 0.0152–0.0076 m2 g−1 , following Liao’s [18] calculating method, i.e. the specific surface of iron powder = 0.762/D m2 g−1 , D represents the iron diameter, reported as m. Iron powder was pretreated with 0.5 mol L−1 HCl until large amounts of gas escaped (about 3–5 min), then washed with degassed deionized water to remove the residual HCl and then dried in vacuum container for further use. The soil was obtained from no.1 Agricultural Experiment Station of Northwest A & F University, China, air-dried and sieved via a 2.0 mm nylon screen. Soil samples used for chemical analysis additionally sieved via a 1.0 mm nylon screen. The properties of the soil can be summarized as: pH (soil: water = 1:2.5 w/v) 8.35 ± 0.12, moisture content 5.21%, organic carbon 1.71%, total N (Kjeldahl-N) 523 mg kg−1 , total iron 3.68%, total copper 29.05 mg kg−1 , nitrate-N 0.94 mg kg−1 , nitriteN 0.018 mg kg−1 , ammonium-N 0.595 mg kg−1 , Cl− 96 mg kg−1 , SO4 2− 75 mg kg−1 , HCO3 − 82 mg kg−1 , dissolvable iron and
Tapered flask (250 mL) was used as the reactor. 42.2 g raw loess soil (Earth-cumuli-Orthic Anthrosoles), corresponding to 40.00 g dry soil (oven dry at 105 ◦ C), and 5.00 g (in excess) pretreated iron powder were added into each reactor, followed by adding nitrate and specific ion stock solutions, resulting in a final ratio of liquid to soil V/W = 5:1 (a national standard of water and soil ratio for the determination of soluble ions in soil), a nitrate concentration of 60 mg N L−1 , a cation concentration of 1.0 mM and 2.0 mM, or a anion concentration of 1.0 mM and 3.0 mM. Then the reactors were sealed tightly with rubber stopple, and placed in a complanate shaker at 200 rpm. The headspace was air. At the same time a control test without iron powder and ions and another control test only containing iron powder were prepared. After shaking 1 min, a little suspension was extracted from the reactor using a syringe, and passed through a 0.45 m fiber filter membrane. The filtrate was collected for pH and EC measurement, which was recorded as the initial values. Subsequently, at the designed time interval about 5.0 ml of suspension was extracted from the reactor with a syringe, followed by filter. The filtrate was collected for pH, EC, nitrate, nitrite, ammonium, and dissolved iron measurement. In order to detect the gas phase constitute in the headspace of reactor after reaction, the reactors were flashed using argon gas (>99.99%) to remove air in the headspace in the beginning of reaction. The gas (N2 ) was collected and analyzed using gas chromatograph after reaction. The NH3 was collected in 0.10 M NaOH solution and measured as NH4 + . All tests were conducted in duplicate and the analyses were finished within 24 h at room temperature of 24 ± 1 ◦ C. 2.3. Analytical methods The scanning electron microscopy (SEM, Hitachi S-450) was used to obtain the microstructure and size information of iron powder. The nitrogen gas in the headspace in reactor after reaction was measured using gas chromatograph (GC-7800, Agilent) equipped thermal conductivity detector. The solution pH and EC were determined using a pHS-3 C model exact pH meter (Lei-ci, China) and a DDS-307 model Electric Conductivity Detector (Lei-ci, China), respectively. Nitrate measurement was achieved by the ultraviolet spectrophotometric method at 220 nm and 275 nm using a UV1102 spectrophotometer [19]. Nitrite was analyzed by the hydrazine reduction method at 540 nm, and ammonium was measured using its reaction with phenol and hypochlorite and catalyzed sodium nitroprusside to form indophenol at 636 nm [19]. The dissolved Fe2+ was determined as phenanthroline method at 510 nm. The total dissolved iron was analyzed as Fe2+ after its reduction using hydroxylamine hydrochloride [19]. All results presented are the average value of duplicate. 3. Results and discussion 3.1. Effect of cations As illustrated in Fig. 2a, acid pretreated Fe0 enhanced nitrate removal. Nitrate removal rate of 69.2% was achieved by pretreated Fe0 at 144 h. By contrast, it was tend to constant after 120 h and only 49.6% removal was observed using raw Fe0 alone at 144 h. It might be because acid pretreated removed the passive layer (e.g., sulfide and oxide (Fe2 O3 )) on the Fe0 surface. A rotten-egg smell was smelled during acid pretreatment. Previous studies also showed
116
C. Tang et al. / Journal of Hazardous Materials 231–232 (2012) 114–119
70
a
blank
acidFe0
50
8
2.0mM Fe2+ 1.0mM Fe3+
30
2.0mM Fe3+
6
1.0mM Cu2
20
4
2.0mM Cu2
10
2
0
Blank
Acid Fe0
1.0 mM Fe2+
2.0 mM Fe2+
1.0 mM Fe3+
2.0 mM Fe3+
1.0 mM Cu2+
2.0 mM Cu2+
0
0
25
50
75
100 125 Time (h)
150
175
200
0
25
50
75
100
125
150
Time (h) 25
6
c
3
Blank
Acid Fe0
1.0 mM Fe2+
2.0 mM Fe2+
1.0 mM Fe3+
2.0 mM Fe3+
1.0 mM Cu2+
2.0 mM Cu2+
2
Acid Fe0
20
mg /L
4
Fe0
d
1 mMFe2+ 2 mMFe2+
15
Am mo nia- N
5
Nitrite-N(mg/ L)
b
10
1.0mM Fe2+
40 Ni trate-N
12
Fe0
pH
mg /L
60
1 mMFe3+ 2 mMFe3+
10
1 mMCu2+ 2 mMCu2+
5
1 0
0
0
50
100 Time (h)
150
0
200
05
100
150
200
Time (h)
Fig. 2. Nitrate reduction by Fe0 in the presence of loess soil and different cation: (a) nitrate nitrogen, (b) pH, (c) nitrite nitrogen, and (d) ammonia nitrogen. Blank treatment was a nitrate solution without Fe0 and additional cation. The same means in the Figs. 3 and 4.
that acid pretreatment enhanced Fe0 reactivity toward nitrate [20] and perchlorate [21] reduction. The introduction of cation significantly promoted nitrate reduction by pretreated Fe0 . 96%, 99%, 92% and 98% of nitrate removal were observed after 144 h in the presence of 1.0 mM Fe2+ , 2.0 mM Fe2+ , 1.0 mM Cu2+ and 2.0 mM Cu2+ , respectively. But complete removal of nitrate was achieved within 30 h and 20 h in the presence of 1.0 mM and 2.0 mM Fe3+ , respectively. The pH in most systems was above 8.0 after 3.0 h due to the release of OH− from iron corrosion (Fig. 2b). The results implied that nitrate reduction by Fe0 could occur at pH > 8.0. Previous studies showed that nitrate reduction by Fe0 was effective under acidic condition (pH < 7.0) but neglectable in alkaline condition (pH > 8.0) [7,8,22]. In the systems including pretreated Fe0 or Fe0 /cation, the added nitrate was removed rapidly in the beginning 20 h and then gradually decreased. It implied that nitrate reduction mainly occurred in the presence of fresh Fe0 . It was well known that Fe0 reaction with pollutants was surface mediated [23]. This was consistent with former researches that the decontamination efficiency of Fe0 was fast in the beginning of reaction or polished Fe0 by ultrasound [24,25]. As the reaction process, iron corrosion products (mainly as hydroxide and iron oxides) and precipitates (e.g., CaCO3 , FeCO3 ) deposited on the surface of Fe0 due to increasing pH, which decreased Fe0 reactivity and restrained the further nitrate reduction [26]. On the other hand, some iron corrosion products such as Fe2+ , Fe(OH)+ , Fe(OH)2 , and green rust have great potential for nitrate reduction [27]. Huang et al. [28] reported that the oxide film on the surface of Fe0 could not prevent the iron from reducing nitrate if plenty of Fe2+ was available with a stoichiometry of 0.75 mol Fe2+ for 1.0 mol nitrate reduction to ammonium under neutral conditions. Once the Fe2+ was depleted, the reaction would stop immediately. Another study also showed that Fe2+ could accelerate iron corrosion and facilitate passive oxides transformation into conductor (Fe3 O4 ), which favored electron transfer and enhanced nitrate reduction by Fe0 [16]. The key role of Fe2+
might occur in this study, because Fe2+ could be produced from redox reaction after external cation was added as below: 2Fe3+ + Fe0 ↓ → 3Fe2+ 2+
Cu
0
0
+ Fe ↓ → Cu ↓ + Fe
(1) 2+
(2)
The added cation enhanced nitrate reduction with an order of Fe3+ > Fe2+ > Cu2+ in the same concentration. Apparently, the direct addition of Fe2+ is better than indirect from Cu2+ replacement (Eq. (2)). Although dissolved Fe2+ was not detectable, Fe(OH)2 or surface adsorbed Fe2+ might contribute the enhancement. Previous studies reported Cu0 could accelerate electron transfer and then increase nitrate reduction by Fe0 as a catalyst [6,14]. But in this study its catalytic efficiency in enhancing electron transfer might be inhibited due to passivation by non-conducting soils and precipitates (e.g., iron oxides, hydroxide). Nitrite accumulated sharply to nearly 6 mg N L−1 in the first 10 h with the effect of Cu2+ (Fig. 2c). More Cu2+ addition favored nitrite accumulation. But nitrite decreased after 16 h and less than 1.0 mg N L−1 in 30 h. An et al. [29] also reported that Fe–Cu nanoparticles significantly enhanced nitrate reduction comparing with nano-Fe alone, but more nitrite was produced. Less than 2.0 mg N L−1 nitrite accumulated in the beginning and subsequently decreased rapidly in the presence of Fe2+ or Fe3+ . It might be because Cu2+ or Cu0 could catalyze nitrate reduction to nitrite but it was weak or not in catalyzing nitrite to ammonia [30], which led to the accumulation of nitrite at the beginning of the reaction. Subsequently, Cu2+ or Cu0 lost its catalytic activity due to precipitation and passivation by nonconducting substances. But the accumulated nitrite disappeared later due to the reduction of Fe0 or corrosion products [27]. In the systems with external addition of cation, electric conductivity (EC) decreased rapidly in the beginning of reaction and then stabilized after 16 h (data not shown). This was due to the consumption of nitrate and adsorption/co-precipitation of ions on the
C. Tang et al. / Journal of Hazardous Materials 231–232 (2012) 114–119 70
A
60
(1.0 mM)
mg/L Nitrate-N
oxalate citrate phosphate
bicarbonate oxalate
30
citrate phosphate
20
acetate
acetate
10
sulfate
40
10
0
0 25
50
75
100
125
150
175
0
25
50
Time(h)
10.5
900
11
10.1
800
10.5
9.7
700 pH
9.3
600
Conductivity 8.9
500
8.5
pH
Conductivity (us/cm)
pH
a
40
100
60 80 Time (h)
pH
600 pH Conductivity
500
8.5 400
8 60 80 Time(h)
1100 900 700
pH 9
Conductivity 500
8.5
300
8 0
Conductivity (us/cm)
9.5
40
b
20
40
60 80 Time (h)
100
d
700
10
20
175
120
10.2
c
0
150
10
120
10.5
9
125
100
120
9.8
750 650
9.4
pH
20
100
9.5
400 0
75
Time(h)
Conductivity (us/cm)
0
550
pH
9
Conductivity 450
8.6
Conductivity (us/cm)
mg/L
bicarbonate
20
Acid Fe0
50
sulfate
30
Blank
chloride
chloride
40
(3.0 mM)
60
AcidFe0
50
Nitrate-N
B
Blank
117
350
8.2 0
20
40
80 60 Time (h)
100
120
Fig. 3. Nitrate reduction by Fe0 as affected by different external anions: (A) 1.0 mM; (B) 3.0 mM, and the pH and EC of 3.0 mM anions: (a) Cl− , (b) SO4 2− , (c) HCO3 − , (d) CH3 COO− .
corrosion products of iron. Finally, both the reduction reaction and adsorption reached equilibrium and then EC kept constant. In all systems containing Fe0 , ammonium was the major final product from nitrate reduction (Fig. 2d). More ammonia was detected in the system with more nitrate removal. The total nitrogen recovery in liquid phase (i.e., NO3 − -N, NO2 − -N and NH4 + -N) and gas phase (NH3 ) was 56–73% (data not shown). It was due to the adsorption of ammonium onto soil and precipitates. NH3 gas was observed in the headspace of reactor after reaction due to relative high pH (>8.0) and stirring. But no N2 and other nitrogen oxide compounds were detected in headspace. Ammonia volatilization in the system with high pH (8–9) was also observed in previous studies [31,32]. Dissolved iron and copper were undetectable in the system because of adsorbed and deposited in high pH. 3.2. Effect of anions As showed in Fig. 3A and B, all anions except PO4 3− enhanced nitrate reduction by Fe0 . Contrarily, PO4 3− inhibited nitrate reduction. The influence was favored at higher concentration. Similar to cation, nitrate decreased sharply during the first 16 h and then slowed down. But it took longer time before they reached equilibrium comparing with cation. In the control treatment (i.e., only pretreated iron without external anion addition) 72.5% of nitrate
removal was observed after 124 h. But it increased to 76.6% and 84.7% in the presence of 1.0 mM and 3.0 mM external addition of Cl− , respectively (Fig. 3A, B). The corresponding increase of ammonium was also observed (data not shown). Nevertheless, insignificant nitrite accumulated and nearly disappeared after 60 h. The enhancement of nitrate reduction by Fe0 might result from the promotion of iron pitting corrosion and the increase of ionic strength by Cl− [33]. Iron corrosion and electron release was the premise of pollutants degradation by Fe0 [23]. Nitrite, as the intermediate product from nitrate reduction, was not stable and easily converted to ammonium by Fe0 and iron corrosion products [27,34]. The nitrate removal efficiency was 84.4% and 92.9% after 124 h in the presence of 1.0 mM and 3.0 mM SO4 2− , respectively. The external addition of ion increased ionic strength and then accelerated iron corrosion. Distinctly, sulfate contributed stronger ion strength than chloride with the same concentration. Hence, its enhancement was better. Moreover, previous studies indicated that the promotion of SO4 2− and Cl− might contribute to the increase of surface reactivity or sorption capacity of Fe0 and the production of green rust (e.g., [Fe4 II Fe2 III (OH)12 SO4 ·yH2 O], GRSO 2− ), which 4
was also effective for nitrate reduction [10,27,35]. It was proved that both GRSO 2− and Fe4.5 Fe1.5 (OH)12 Cl1.5 ·xH2 O (GRCl− ) could sto4
ichiometrically reduce nitrate to ammonium [27]. Thus, in the system externally adding Cl− and SO4 2− , nitrate reduction was
C. Tang et al. / Journal of Hazardous Materials 231–232 (2012) 114–119
10.2
650
Conductivity (us/cm)
10
pH
9.8 9.6
pH
9.4
Conductivity
550
9.2 9
b 10.5
700
10
650
9.5
600
pH
a
9
450 0
20
40
60
80
100
550 pH Conductivity
8.5
120
500 0
20
40
Time(h)
c
Conductivity (us/cm)
118
60
80
100
120
Time(h)
10
pH
1000 9
900 pH Conductivity
8.5
800
Conductivity (us/cm)
1100 9.5
700
8 0
20
40
60
80
100
120
Time(h) Fig. 4. The time course of pH and EC in the presence of 3.0 mM anion: (a) PO4 3− , (b) C2 O4 2− , (c) citrate.
probably partially resulted from green rusts (GRCl− ) and GRSO
4
2−
.
The promotion of HCO3 − was similar to Cl− at the same concentration, resulting in a nitrate removal of 76.4% and 85.7% in the presence of 1.0 mM and 3.0 mM, respectively. Similar to GRSO 2− 4
and GRCl− , the GRHCO − also had potential for nitrate reduction 3 [36]. A green color was observed after settling during experiments in the system with externally adding Cl− , SO4 2− and HCO3 − . The introduction of 1.0 mM and 3.0 mM CH3 COO− achieved 88.25% and 93.4% nitrate removal, respectively. As a whole, the positive effects of the four anions above decreased in the order: CH3 COO− > SO4 2− > Cl− ≈ HCO3 − . Although the first ion was organic and the others were inorganic, they exhibited similar variation in pH and EC during the whole reaction period. The solution pH increased and EC decreased significantly within the first 40 h, and then kept stable (Fig. 3a, b, c, d). Both alkaline release from iron corrosion and acidity consumption from nitrate reduction by Fe0 contributed to pH increase. Nitrate reduction and subsequent ammonium adsorption, the adsorption and co-precipitation of other ions onto iron hydroxides all contributed to EC decrease. It implied that other co-existent contaminants (e.g., heavy metals) would be removed simultaneously by adsorption and precipitation. The nitrate removal rate was 62.4% and 60.9% (Fig. 3A, B) in the system externally adding 1.0 mM and 3.0 mM PO4 3− , respectively. The nitrogen recovery was the highest in all anions (data not shown). Phosphate has been reported to form co-precipitation and inner-sphere complexes on the surface of iron inhibited iron corrosion and electron transfer [37,38]. Although the addition of phosphate could increase ion strength and promote iron corrosion, the competitive adsorption was stronger. Oxalate achieved similar nitrate removal rate as Cl− in the same concentration (76.3% and 84.2%) (Fig. 3A, B). Among all anions, citrate achieved the best promotion to nitrate reduction by Fe0 . Almost complete removal of nitrate (>95%) was observed after 124 h in the presence of 1.0 mM and 3.0 mM citrate. The reactors containing citrate were very clear at the last period of the experiment, though they were turbid in the beginning of reaction as other systems. But the reactors containing
other anions were always turbid. Considering citrate was usually used as adsorbent or chelator in industry, it might have the same effect in this study. Citrate as a chelator could destabilize and finally dissolve the passive oxide layer on the surface of iron [37]. Although PO4 3− , oxalate and citrate behaved differently on nitrate reduction, they exhibited similar variation in pH and EC. The solution pH and EC varied repetitive and showed opposite to each other (Fig. 4). A possible explanation was that the added ion behaved as a buffer alone or together with some ions in soil solution, and thus they prevented the solution pH from fluctuating. The low pH led to dissolve the precipitates and/or desorb the adsorbed ions, which caused a rise of EC. Contrarily, the high pH induced precipitation/co-precipitation and then more adsorption, which decreased the EC. Consequently, the solution pH and EC changed simultaneously and exhibited a fluctuant situation, especially with higher concentration. In conclusion, all anions excluding PO4 3− revealed positive effects on nitrate reduction by Fe0 in the presence of loess soil. The promotion effect decreased in the order: citrate > acetate > SO4 2− > Cl− ≈ HCO3 − ≈ oxalate PO4 3− . Similar to cation, nitrogen recovery was only 61–78%. Some nitrite accumulation was observed in the beginning, but less than 1.0 mg N L−1 was detected after reaction in all systems with different anion present. 4. Fe0 -based PRB implication in the loess plateau area According to the above results, reductive removal of nitrate by Fe0 from an alkaline (pH > 8) soil was possible. The major anions (Cl− , SO4 2− , HCO3 − ) can accelerate nitrate reduction due to the promotion of iron corrosion and the possible occurrence of green rusts. The Fe0 corrosion products, Fe(II) and Fe(III), significantly enhanced nitrate reduction. Moreover, nitrate can be reduced to nitrogen gas by hydrogenotrophic denitrification process which consumed H2 deriving from iron corrosion [39,40]. More iron powder may be consumed in the regions where the concentration of PO4 3− is higher. Some organic ligands (e.g.,
C. Tang et al. / Journal of Hazardous Materials 231–232 (2012) 114–119
citrate, acetate and oxalate) would promote not only nitrate chemical reduction by Fe0 but also microorganism denitrification. For example, the denitrification rates decreased in the of order acetate > H2 > S > thiosulphate > ferrous iron in microcosm [41]. Certainly, there are lots of concomitant contaminants in the specific environment, of which the effects still need to investigate in detail. The iron in reactors after reaction was black and agglomeration. The passivation of iron surface resulting from high pH should account for the reaction expiration, which would decrease the longevity of Fe0 [42]. Thus, in an alkaline loess plateau area, more Fe0 consumption and relative short longevity of PRB may be the main concern. But it is still an alternative approach to remediate soil and groundwater nitrate contamination if no other better technologies to be developed in future. 5. Conclusions The results indicated that nitrate reduction by Fe0 in the alkaline loess plateau area was possible. High pH did not inhibit nitrate reduction reaction immediately. By contrast, the remove efficiency could be accelerated by some cations (Fe3+ , Cu2+ and Fe2+ ) and anions (citrate, acetate, oxalate, Cl− , SO4 2− and HCO3 − ). At the same concentration, the promotion effect of the cation decreased in the order of Fe3+ > Fe2+ > Cu2+ , but the anions decreased in this order of citrate > acetate > SO4 2− > Cl− ≈ HCO3 − ≈ oxalate PO4 3− . Only PO4 3− exhibited an inhibition for nitrate reduction by Fe0 . Ammonium was the major detectable final product from nitrate reduction. Nitrite, the intermediate product of nitrate reduction, was less than 1.0 mg N L−1 after 100 h reaction. But the nitrogen recovery in liquid and gas phase was only 56–78% in all systems. Dissolved iron was not detectable. The pH and EC varied depending on the character of added ion. This study implied that it is a potential technique to use Fe0 -based PRB for in situ remediation of nitrate contamination in the loess plateau area. Acknowledgements We acknowledge Mr. Su Chunming, National Risk Management Research Lab., U. S. Environmental Protection Agency, 919 Kerr Research Drive, Ada, Oklahoma, for editing the paper, and Dr. Brian D. Strahm, University of Washington, College of Forest Resources, Seattle, U.S., for correcting expression in the paper. We also thank the professor Yao Yaqin for SEM micrograph, College of Life Science, Northwest A & F University. We also acknowledge editor and anonymous reviewers for their suggestion and comments. References [1] Y. Zhang, Y.X. Chen, H.Y. Liu, Countermeasure and removal of nitrate in groundwater, Agro-Environ. Prot. 21 (2) (2002) 183–184 (in Chinese). [2] G.H. Jiang, W.K. Wang, Analysis of nitrate pollution of groundwater of Guanzhong basin and countermeasure, Water Resour. Prot. 68 (2) (2002) 6–8 (in Chinese). [3] J.O. Lundberg, E. Weitzberg, J.A. Cole, N. Benjamin, Nitrate, bacteria and human health, Nat. Rev. Microbiol. 2 (2004) 593–602. [4] S.M. Li, Survey and analysis on the main chemical and physical characters of water in Hongsipu irrigated area of Ningxia, J. Ningxia Agric. Coll. 24 (3) (2003) 37–39 (in Chinese). [5] X. Fan, X. Guan, J. Ma, H. Ai, Kinetics and corrosion products of aqueous nitrate reduction by iron powder without reaction conditions control, J. Environ. Sci. 21 (2009) 1028–1035. [6] S.M. Hosseini, B. Ataie-Ashtiani, M. Kholghi, Nitrate reduction by nano-Fe/Cu particles in packed column, Desalination 276 (2011) 214–221. [7] C.P. Huang, H.W. Wang, P.C. Chiu, Nitrate reduction by metallic iron, Water Res. 32 (8) (1998) 2257–2264. [8] Y.H. Huang, T.C. Zhang, Effects of low pH on nitrate reduction by iron powder, Water Res. 38 (11) (2004) 2631–2642. [9] D.W. Blowes, C.J. Ptacek, S.G. Benner, C.W.T. McRae, T.A. Bennett, R.W. Puls, Treatment of inorganic contaminants using permeable reactive barriers, J. Contam. Hydrol. 45 (2000) 123–137.
119
[10] J.F. Devlin, K.O. Allin, Major anion effects on the kinetics and reactivity of granular iron in glass-encased magnet batch reactor experiments, Environ. Sci. Technol. 39 (6) (2005) 1868–1874. [11] L. Xie, C. Shang, The effects of operational parameters and common anions on the reactivity of zero-valent iron in bromate reduction, Chemosphere 66 (2007) 1652–1659. [12] Y. Liu, T. Phenrat, G.V. Lowry, Effect of TCE concentration and dissolved groundwater solutes on nZVI-promoted TCE dechlorination and H2 evolution, Environ. Sci. Technol. 41 (2007) 7881–7887. [13] T. Liu, I.M.C. Lo, Influences of humic acid on Cr(VI) removal by zero-valent iron from groundwater with various constituents: implication for long-term PRB, Water Air Soil Pollut. 216 (2011) 473–483. [14] Y.H. Shih, M.Y. Chen, Y.F. Su, Pentachlorophenol reduction by Pd/Fe bimetallic nanoparticles: effects of copper, nickel, and ferric cations, Appl. Catal. B: Environ. 105 (2011) 24–29. [15] M. Hou, H. Wan, T. Liu, Y. Fan, X. Liu, X. Wang, The effect of different divalent cations on the reduction of hexavalent chromium by zerovalent iron, Appl. Catalysis B: Environ. 84 (2008) 170–175. [16] J. Xu, Z. Hao, C. Xie, X. Lv, Y. Yang, X. Xu, Promotion effect of Fe2+ and Fe3 O4 on nitrate reduction using zero-valent iron, Desalination 284 (2012) 9–13. [17] Y.H. Huang, T.C. Zhang, Enhancement of nitrate reduction in Fe0 -packed columns by selected cations, J. Environ. Eng. 131 (7) (2005) 603–611. [18] C.H. Liao, S.F. Kang, Y.W. Hsu, Zero-valent iron reduction of nitrate in the presence of ultraviolet light, organic matter and hydrogen peroxide, Water Res. 37 (2003) 4109–4118. [19] EPA, American Public Health Association, standard methods for the examination of water and wastewater, 19th, 1995. [20] S.F. Cheng, C.Y. Huang, J.Y. Liu, Study of different methods for enhancing the nitrate removal efficiency of a zero-valent metal process, Water Sci. Technol. 53 (2006) 81–87. [21] H. Huang, G.A. Sorial, Perchlorate remediation in aquatic systems by zerovalent iron, Environ. Eng. Sci. 24 (2007) 917–926. [22] C. Seunghee, M.L. Howand, Nitrate reduction by zero-valent iron under different pH regimes, Appl. Geochem. 19 (2004) 335–342. [23] C. Noubactep, A critical review on the process of contaminant removal in Fe0 –H2 O systems, Environ. Technol. 29 (2008) 909–920. [24] Y.J. Tsai, F.C. Chou, T.C. Cheng, Coupled acidification and ultrasound with iron enhances nitrate reduction, J. Hazard. Mater. 163 (2009) 743–747. [25] Q.J. Rasheed, K. Pandian, K. Muthukumar, Treatment of petroleum refinery wastewater by ultrasound-dispersed nanoscale zero-valent iron particles, Ultrason. Sonochem. 18 (2011) 1138–1142. [26] W. Kamolpornwijit, L.Y. Liang, G.R. Moline, T. Hart, O.R. West, Identification and quantification of mineral precipitation in Fe0 filings from a column study, Environ. Sci. Technol. 38 (2004) 5757–5765. [27] H.C.B. Hansen, G. Susanne, E. Marianne, Kinetics of nitrate reduction by green rust—effects of interlayer anion and Fe(II): Fe(III) ratio, Appl. Clay Sci. 18 (2001) 81–91. [28] Y.H. Huang, T.C. Zhang, J.S. Patrick, Effects of oxide coating and selected cations on nitrate reduction by iron metal, J. Environ. Qual. 32 (2003) 1306–1315. [29] Y. An, T. Li, Z. Jin, M. Dong, H. Xia, W. Xue, Effect of bimetallic and polymercoated Fe nanoparticles on biological denitrification, Bioresour. Technol. 101 (2010) 9825–9828. [30] Z.W. Hao, X.H. Xu, J. Jin, Simultaneous removal of nitrate and heavy metals by iron metal, J. Zhejiang Univ. Sci. B 5 (2005) 307–310. [31] Y.H. Hwang, D.G. Kim, Y.T. Ahn, C.M. Moon, H. Shin, Fate of nitrogen species in nitrate reduction by nanoscale zero valent iron and characterization of the reaction kinetics, Water Sci. Technol. 61 (2010) 705–712. [32] Y.H. Hwang, D.G. Kim, H.S. Shin, Mechanism study of nitrate reduction by nano zero valent iron, J. Hazard. Mater. 185 (2011) 1513–1521. [33] T. Fukuda, T. Mizuno, The evaluation of pitting corrosion from the spectrum slope of noise fluctuation on iron and 304 stainless steel electrodes, Corros. Sci. 38 (1996) 1085–1091. [34] H.Y. Hu, N. Goto, K. Fujie, Effect of pH on the reduction of nitrite in water by metallic iron, Water Res. 35 (2001) 2789–2793. [35] C.B.H. Hans, B.K. Christian, N.K. Hanne, Abiotic nitrate reduction to ammonium: key role of green rust, Environ. Sci. Technol. 30 (1996) 2053–2056. [36] C.M. Su, R.W. Puls, Significance of iron (II, III) hydroxycarbonate green rust in arsenic remediation using zerovalent him in laboratory column tests, Environ, Sci. Technol. 38 (2004) 5224–5231. [37] H. Harms, H.P. Volkland, G. Repphun, A. Hiltpolt, O. Wanner, A.J.B. Zehnder, Action of chelators on solid iron in phosphate-containing aqueous solution, Corros. Sci. 45 (2003) 1717–1732. [38] C.M. Su, R.W. Puls, Nitrate reduction by zerovalent iron: Effects of formate, oxalate, citrate, chloride, sulfate, borate and phosphate, Environ. Sci. Technol. 38 (2004) 2715–2720. [39] Y. An, T. Li, Z. Jin, M. Dong, Q. Li, Nitrate degradation and kinetic analysis of the denitrification system composed of iron nanoparticles and hydrogenotrophic bacteria, Desalination 252 (2010) 71–74. [40] S. Biswas, P. Bose, Zero-valent iron-assisted autotrophic denitrification, J. Environ. Eng. 131 (2005) 1212–1220. [41] J.F. Devlin, R. Eedy, B.J. Butler, The effects of electron donor and granular iron on nitrate transformation rates in sediments from a municipal water supply aquifer, J. Contam. Hydrol. 46 (2000) 81–97. [42] T. Kohn, K.J.T. Livi, A.L. Roberts, P. Vikesland, Longevity of granular iron in groundwater treatment processes: corrosion product development, Environ. Sci. Technol. 39 (2005) 2867–2879.
Contents lists available at SciVerse ScienceDirect
Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat
Effect of common ions on nitrate removal by zero-valent iron from alkaline soil Cilai Tang, Zengqiang Zhang ∗ , Xining Sun College of Resources and Environment, Northwest A&F University, Yangling, Shannxi 712100, China
h i g h l i g h t s The selected cations and anions enhance nitrate reduction by Fe0 in alkaline soil. Ammonium is major final product from nitrate reduction. The results prove the feasibility of using Fe0 for groundwater nitrate remediation.
a r t i c l e
i n f o
Article history: Received 29 March 2012 Received in revised form 21 June 2012 Accepted 21 June 2012 Available online 29 June 2012 Keywords: Zero-valent iron Nitrate reduction Soil and groundwater remediation Loess soil Cations Anions
a b s t r a c t Zero-valent iron (Fe0 )-based permeable reactive barrier (PRB) technology has been proved to be effective for soil and groundwater nitrate remediation under acidic or near neutral conditions. But few studies have been reported about it and the effects of coexistent ions under alkaline conditions. In this study, nitrate reduction by Fe0 was evaluated via batch tests in the presence of alkaline soil and common cation (Fe2+ , Fe3+ and Cu2+ ) and anion (citrate, oxalate, acetate, SO4 2− , PO4 3− , Cl− and HCO3 − ). The results showed that cation significantly enhanced nitrate reduction with an order of Fe3+ > Fe2+ > Cu2+ due to providing Fe2+ directly or indirectly. Most anions enhanced nitrate reduction, but PO4 3− behaved inhibition. The promotion decreased in the order of citrate > acetate > SO4 2− > Cl− ≈ HCO3 − ≈ oxalate PO4 3− . Ammonium was the major final product from nitrate reduction by Fe0 , while a little nitrite accumulated in the beginning of reaction. The nitrogen recovery in liquid and gas phase was only 56–78% after reaction due to ammonium adsorption onto soil. The solution pH and electric conductivity (EC) varied depending on the specific ion added. The results implied that PRB based Fe0 is a potential approach for in situ remediation of soil and groundwater nitrate contamination in the alkaline conditions. © 2012 Elsevier B.V. All rights reserved.
1. Introduction The loess plateau, an arid or semi-arid region, is the predominant terrain in the northwest of China, comprising an area of 6.24 × 105 km2 . Loess soil is an alkaline soil. In this area, water resource is scarce and groundwater is the predominant water source for people living, agricultural and industrial use. But as the industrial and agricultural development, nitrogen-containing industrial waste effluent discharge, excessive use of nitrogen fertilizer to improve food output for increasing population induced increasingly nitrate contamination in the groundwater [1]. An investigation on sub-aqueous nitrate contamination in Guanzhong basin (a typical loess plateau), Shaanxi Province in China, indicated that nitrate concentration in most groundwater markedly exceeded the drinking water standard (GB 5749-2006) of China (10 mg N L−1 ) [2]. Excessive nitrate in drinking water can cause cancer and other diseases [3]. Loess soil is an alkaline soil that has an average pH
∗ Corresponding author. E-mail address: [email protected] (Z. Zhang). 0304-3894/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jhazmat.2012.06.042
value above 8.0, and HCO3 − , SO4 2− , Cl− , Na+ , Ca2+ , and K+ are the major ions in the soil and groundwater [4]. Many studies have shown that nitrate reduction by Fe0 was an acid-driven and surface-mediated process, which was spontaneous under acidic condition, or near neutral condition with the help of catalysts [5–8]. Permeable reactive barrier (PRB) based Fe0 has been successfully used for in situ groundwater contamination remediation, including nitrate, under acidic or near neutral conditions [9]. But no study has been reported whether or not it could work under alkaline condition, such as in alkaline loess plateau. Granular iron reactivity is predominantly controlled by the groundwater geochemistry (e.g., coexistent ion and pH) [10,11]. The coexistent ions behaved different impacts on Fe0 reactivity depending on the target pollutants [12,13]. Fe2+ , Fe3+ and corresponding hydroxides are the oxidative products from Fe0 corrosion. Lots of previous studies showed that Fe2+ or Fe3+ enhanced Fe0 degradation of pentachlorophenol [14], chromate [15] and nitrate [16,17]. Cu(II) was usually added as a catalyst for enhancing pollutants removal by Fe0 [6,14]. Cations (Fe2+ , Fe3+ and Cu2+ ) without soil have been proved to promote nitrate reduction by Fe0 but there were no distinct effect by Ca2+ , Na+ and K+ based on the former studies [17]. Therefore, we
C. Tang et al. / Journal of Hazardous Materials 231–232 (2012) 114–119
115
copper were not detectable. All processes were performed as the methods for the examination of water and wastewater of China (4th edition). 2.2. Experiments
Fig. 1. SEM photograph of iron grains.
also investigated the effects of Fe2+ , Fe3+ and Cu2+ on nitrate reduction by Fe0 in the presence of loess soil. Moreover, the common inorganic ions such as Cl− , HCO3 − , SO4 2− , PO4 3− and potential coexistent organic ions such as citrate, oxalate and acetate have been selected to evaluate their effects on nitrate removal by Fe0 . The objective of this study was to evaluate the effects of common ions in soil and groundwater on nitrate reduction by Fe0 in the presence of alkaline soil, aiming to give some possible guidance to practical application of PRB technology for in situ removal of nitrate in the loess plateau area in the future. 2. Materials and methods 2.1. Materials Unless otherwise indicated, all chemicals used were reagent grade and all aqueous solutions were prepared with deionized water. Nitrate solution was prepared using NaNO3 . Cation and anion stock solutions were prepared with corresponding chloride and sodium salts, respectively. The iron particle (Tianjin Zonghengxing Industrial &Trading Chemical Reagent Co., China) size was approximately 50–100 m in diameter (Fig. 1), irregular in shape, with a specific surface area 0.0152–0.0076 m2 g−1 , following Liao’s [18] calculating method, i.e. the specific surface of iron powder = 0.762/D m2 g−1 , D represents the iron diameter, reported as m. Iron powder was pretreated with 0.5 mol L−1 HCl until large amounts of gas escaped (about 3–5 min), then washed with degassed deionized water to remove the residual HCl and then dried in vacuum container for further use. The soil was obtained from no.1 Agricultural Experiment Station of Northwest A & F University, China, air-dried and sieved via a 2.0 mm nylon screen. Soil samples used for chemical analysis additionally sieved via a 1.0 mm nylon screen. The properties of the soil can be summarized as: pH (soil: water = 1:2.5 w/v) 8.35 ± 0.12, moisture content 5.21%, organic carbon 1.71%, total N (Kjeldahl-N) 523 mg kg−1 , total iron 3.68%, total copper 29.05 mg kg−1 , nitrate-N 0.94 mg kg−1 , nitriteN 0.018 mg kg−1 , ammonium-N 0.595 mg kg−1 , Cl− 96 mg kg−1 , SO4 2− 75 mg kg−1 , HCO3 − 82 mg kg−1 , dissolvable iron and
Tapered flask (250 mL) was used as the reactor. 42.2 g raw loess soil (Earth-cumuli-Orthic Anthrosoles), corresponding to 40.00 g dry soil (oven dry at 105 ◦ C), and 5.00 g (in excess) pretreated iron powder were added into each reactor, followed by adding nitrate and specific ion stock solutions, resulting in a final ratio of liquid to soil V/W = 5:1 (a national standard of water and soil ratio for the determination of soluble ions in soil), a nitrate concentration of 60 mg N L−1 , a cation concentration of 1.0 mM and 2.0 mM, or a anion concentration of 1.0 mM and 3.0 mM. Then the reactors were sealed tightly with rubber stopple, and placed in a complanate shaker at 200 rpm. The headspace was air. At the same time a control test without iron powder and ions and another control test only containing iron powder were prepared. After shaking 1 min, a little suspension was extracted from the reactor using a syringe, and passed through a 0.45 m fiber filter membrane. The filtrate was collected for pH and EC measurement, which was recorded as the initial values. Subsequently, at the designed time interval about 5.0 ml of suspension was extracted from the reactor with a syringe, followed by filter. The filtrate was collected for pH, EC, nitrate, nitrite, ammonium, and dissolved iron measurement. In order to detect the gas phase constitute in the headspace of reactor after reaction, the reactors were flashed using argon gas (>99.99%) to remove air in the headspace in the beginning of reaction. The gas (N2 ) was collected and analyzed using gas chromatograph after reaction. The NH3 was collected in 0.10 M NaOH solution and measured as NH4 + . All tests were conducted in duplicate and the analyses were finished within 24 h at room temperature of 24 ± 1 ◦ C. 2.3. Analytical methods The scanning electron microscopy (SEM, Hitachi S-450) was used to obtain the microstructure and size information of iron powder. The nitrogen gas in the headspace in reactor after reaction was measured using gas chromatograph (GC-7800, Agilent) equipped thermal conductivity detector. The solution pH and EC were determined using a pHS-3 C model exact pH meter (Lei-ci, China) and a DDS-307 model Electric Conductivity Detector (Lei-ci, China), respectively. Nitrate measurement was achieved by the ultraviolet spectrophotometric method at 220 nm and 275 nm using a UV1102 spectrophotometer [19]. Nitrite was analyzed by the hydrazine reduction method at 540 nm, and ammonium was measured using its reaction with phenol and hypochlorite and catalyzed sodium nitroprusside to form indophenol at 636 nm [19]. The dissolved Fe2+ was determined as phenanthroline method at 510 nm. The total dissolved iron was analyzed as Fe2+ after its reduction using hydroxylamine hydrochloride [19]. All results presented are the average value of duplicate. 3. Results and discussion 3.1. Effect of cations As illustrated in Fig. 2a, acid pretreated Fe0 enhanced nitrate removal. Nitrate removal rate of 69.2% was achieved by pretreated Fe0 at 144 h. By contrast, it was tend to constant after 120 h and only 49.6% removal was observed using raw Fe0 alone at 144 h. It might be because acid pretreated removed the passive layer (e.g., sulfide and oxide (Fe2 O3 )) on the Fe0 surface. A rotten-egg smell was smelled during acid pretreatment. Previous studies also showed
116
C. Tang et al. / Journal of Hazardous Materials 231–232 (2012) 114–119
70
a
blank
acidFe0
50
8
2.0mM Fe2+ 1.0mM Fe3+
30
2.0mM Fe3+
6
1.0mM Cu2
20
4
2.0mM Cu2
10
2
0
Blank
Acid Fe0
1.0 mM Fe2+
2.0 mM Fe2+
1.0 mM Fe3+
2.0 mM Fe3+
1.0 mM Cu2+
2.0 mM Cu2+
0
0
25
50
75
100 125 Time (h)
150
175
200
0
25
50
75
100
125
150
Time (h) 25
6
c
3
Blank
Acid Fe0
1.0 mM Fe2+
2.0 mM Fe2+
1.0 mM Fe3+
2.0 mM Fe3+
1.0 mM Cu2+
2.0 mM Cu2+
2
Acid Fe0
20
mg /L
4
Fe0
d
1 mMFe2+ 2 mMFe2+
15
Am mo nia- N
5
Nitrite-N(mg/ L)
b
10
1.0mM Fe2+
40 Ni trate-N
12
Fe0
pH
mg /L
60
1 mMFe3+ 2 mMFe3+
10
1 mMCu2+ 2 mMCu2+
5
1 0
0
0
50
100 Time (h)
150
0
200
05
100
150
200
Time (h)
Fig. 2. Nitrate reduction by Fe0 in the presence of loess soil and different cation: (a) nitrate nitrogen, (b) pH, (c) nitrite nitrogen, and (d) ammonia nitrogen. Blank treatment was a nitrate solution without Fe0 and additional cation. The same means in the Figs. 3 and 4.
that acid pretreatment enhanced Fe0 reactivity toward nitrate [20] and perchlorate [21] reduction. The introduction of cation significantly promoted nitrate reduction by pretreated Fe0 . 96%, 99%, 92% and 98% of nitrate removal were observed after 144 h in the presence of 1.0 mM Fe2+ , 2.0 mM Fe2+ , 1.0 mM Cu2+ and 2.0 mM Cu2+ , respectively. But complete removal of nitrate was achieved within 30 h and 20 h in the presence of 1.0 mM and 2.0 mM Fe3+ , respectively. The pH in most systems was above 8.0 after 3.0 h due to the release of OH− from iron corrosion (Fig. 2b). The results implied that nitrate reduction by Fe0 could occur at pH > 8.0. Previous studies showed that nitrate reduction by Fe0 was effective under acidic condition (pH < 7.0) but neglectable in alkaline condition (pH > 8.0) [7,8,22]. In the systems including pretreated Fe0 or Fe0 /cation, the added nitrate was removed rapidly in the beginning 20 h and then gradually decreased. It implied that nitrate reduction mainly occurred in the presence of fresh Fe0 . It was well known that Fe0 reaction with pollutants was surface mediated [23]. This was consistent with former researches that the decontamination efficiency of Fe0 was fast in the beginning of reaction or polished Fe0 by ultrasound [24,25]. As the reaction process, iron corrosion products (mainly as hydroxide and iron oxides) and precipitates (e.g., CaCO3 , FeCO3 ) deposited on the surface of Fe0 due to increasing pH, which decreased Fe0 reactivity and restrained the further nitrate reduction [26]. On the other hand, some iron corrosion products such as Fe2+ , Fe(OH)+ , Fe(OH)2 , and green rust have great potential for nitrate reduction [27]. Huang et al. [28] reported that the oxide film on the surface of Fe0 could not prevent the iron from reducing nitrate if plenty of Fe2+ was available with a stoichiometry of 0.75 mol Fe2+ for 1.0 mol nitrate reduction to ammonium under neutral conditions. Once the Fe2+ was depleted, the reaction would stop immediately. Another study also showed that Fe2+ could accelerate iron corrosion and facilitate passive oxides transformation into conductor (Fe3 O4 ), which favored electron transfer and enhanced nitrate reduction by Fe0 [16]. The key role of Fe2+
might occur in this study, because Fe2+ could be produced from redox reaction after external cation was added as below: 2Fe3+ + Fe0 ↓ → 3Fe2+ 2+
Cu
0
0
+ Fe ↓ → Cu ↓ + Fe
(1) 2+
(2)
The added cation enhanced nitrate reduction with an order of Fe3+ > Fe2+ > Cu2+ in the same concentration. Apparently, the direct addition of Fe2+ is better than indirect from Cu2+ replacement (Eq. (2)). Although dissolved Fe2+ was not detectable, Fe(OH)2 or surface adsorbed Fe2+ might contribute the enhancement. Previous studies reported Cu0 could accelerate electron transfer and then increase nitrate reduction by Fe0 as a catalyst [6,14]. But in this study its catalytic efficiency in enhancing electron transfer might be inhibited due to passivation by non-conducting soils and precipitates (e.g., iron oxides, hydroxide). Nitrite accumulated sharply to nearly 6 mg N L−1 in the first 10 h with the effect of Cu2+ (Fig. 2c). More Cu2+ addition favored nitrite accumulation. But nitrite decreased after 16 h and less than 1.0 mg N L−1 in 30 h. An et al. [29] also reported that Fe–Cu nanoparticles significantly enhanced nitrate reduction comparing with nano-Fe alone, but more nitrite was produced. Less than 2.0 mg N L−1 nitrite accumulated in the beginning and subsequently decreased rapidly in the presence of Fe2+ or Fe3+ . It might be because Cu2+ or Cu0 could catalyze nitrate reduction to nitrite but it was weak or not in catalyzing nitrite to ammonia [30], which led to the accumulation of nitrite at the beginning of the reaction. Subsequently, Cu2+ or Cu0 lost its catalytic activity due to precipitation and passivation by nonconducting substances. But the accumulated nitrite disappeared later due to the reduction of Fe0 or corrosion products [27]. In the systems with external addition of cation, electric conductivity (EC) decreased rapidly in the beginning of reaction and then stabilized after 16 h (data not shown). This was due to the consumption of nitrate and adsorption/co-precipitation of ions on the
C. Tang et al. / Journal of Hazardous Materials 231–232 (2012) 114–119 70
A
60
(1.0 mM)
mg/L Nitrate-N
oxalate citrate phosphate
bicarbonate oxalate
30
citrate phosphate
20
acetate
acetate
10
sulfate
40
10
0
0 25
50
75
100
125
150
175
0
25
50
Time(h)
10.5
900
11
10.1
800
10.5
9.7
700 pH
9.3
600
Conductivity 8.9
500
8.5
pH
Conductivity (us/cm)
pH
a
40
100
60 80 Time (h)
pH
600 pH Conductivity
500
8.5 400
8 60 80 Time(h)
1100 900 700
pH 9
Conductivity 500
8.5
300
8 0
Conductivity (us/cm)
9.5
40
b
20
40
60 80 Time (h)
100
d
700
10
20
175
120
10.2
c
0
150
10
120
10.5
9
125
100
120
9.8
750 650
9.4
pH
20
100
9.5
400 0
75
Time(h)
Conductivity (us/cm)
0
550
pH
9
Conductivity 450
8.6
Conductivity (us/cm)
mg/L
bicarbonate
20
Acid Fe0
50
sulfate
30
Blank
chloride
chloride
40
(3.0 mM)
60
AcidFe0
50
Nitrate-N
B
Blank
117
350
8.2 0
20
40
80 60 Time (h)
100
120
Fig. 3. Nitrate reduction by Fe0 as affected by different external anions: (A) 1.0 mM; (B) 3.0 mM, and the pH and EC of 3.0 mM anions: (a) Cl− , (b) SO4 2− , (c) HCO3 − , (d) CH3 COO− .
corrosion products of iron. Finally, both the reduction reaction and adsorption reached equilibrium and then EC kept constant. In all systems containing Fe0 , ammonium was the major final product from nitrate reduction (Fig. 2d). More ammonia was detected in the system with more nitrate removal. The total nitrogen recovery in liquid phase (i.e., NO3 − -N, NO2 − -N and NH4 + -N) and gas phase (NH3 ) was 56–73% (data not shown). It was due to the adsorption of ammonium onto soil and precipitates. NH3 gas was observed in the headspace of reactor after reaction due to relative high pH (>8.0) and stirring. But no N2 and other nitrogen oxide compounds were detected in headspace. Ammonia volatilization in the system with high pH (8–9) was also observed in previous studies [31,32]. Dissolved iron and copper were undetectable in the system because of adsorbed and deposited in high pH. 3.2. Effect of anions As showed in Fig. 3A and B, all anions except PO4 3− enhanced nitrate reduction by Fe0 . Contrarily, PO4 3− inhibited nitrate reduction. The influence was favored at higher concentration. Similar to cation, nitrate decreased sharply during the first 16 h and then slowed down. But it took longer time before they reached equilibrium comparing with cation. In the control treatment (i.e., only pretreated iron without external anion addition) 72.5% of nitrate
removal was observed after 124 h. But it increased to 76.6% and 84.7% in the presence of 1.0 mM and 3.0 mM external addition of Cl− , respectively (Fig. 3A, B). The corresponding increase of ammonium was also observed (data not shown). Nevertheless, insignificant nitrite accumulated and nearly disappeared after 60 h. The enhancement of nitrate reduction by Fe0 might result from the promotion of iron pitting corrosion and the increase of ionic strength by Cl− [33]. Iron corrosion and electron release was the premise of pollutants degradation by Fe0 [23]. Nitrite, as the intermediate product from nitrate reduction, was not stable and easily converted to ammonium by Fe0 and iron corrosion products [27,34]. The nitrate removal efficiency was 84.4% and 92.9% after 124 h in the presence of 1.0 mM and 3.0 mM SO4 2− , respectively. The external addition of ion increased ionic strength and then accelerated iron corrosion. Distinctly, sulfate contributed stronger ion strength than chloride with the same concentration. Hence, its enhancement was better. Moreover, previous studies indicated that the promotion of SO4 2− and Cl− might contribute to the increase of surface reactivity or sorption capacity of Fe0 and the production of green rust (e.g., [Fe4 II Fe2 III (OH)12 SO4 ·yH2 O], GRSO 2− ), which 4
was also effective for nitrate reduction [10,27,35]. It was proved that both GRSO 2− and Fe4.5 Fe1.5 (OH)12 Cl1.5 ·xH2 O (GRCl− ) could sto4
ichiometrically reduce nitrate to ammonium [27]. Thus, in the system externally adding Cl− and SO4 2− , nitrate reduction was
C. Tang et al. / Journal of Hazardous Materials 231–232 (2012) 114–119
10.2
650
Conductivity (us/cm)
10
pH
9.8 9.6
pH
9.4
Conductivity
550
9.2 9
b 10.5
700
10
650
9.5
600
pH
a
9
450 0
20
40
60
80
100
550 pH Conductivity
8.5
120
500 0
20
40
Time(h)
c
Conductivity (us/cm)
118
60
80
100
120
Time(h)
10
pH
1000 9
900 pH Conductivity
8.5
800
Conductivity (us/cm)
1100 9.5
700
8 0
20
40
60
80
100
120
Time(h) Fig. 4. The time course of pH and EC in the presence of 3.0 mM anion: (a) PO4 3− , (b) C2 O4 2− , (c) citrate.
probably partially resulted from green rusts (GRCl− ) and GRSO
4
2−
.
The promotion of HCO3 − was similar to Cl− at the same concentration, resulting in a nitrate removal of 76.4% and 85.7% in the presence of 1.0 mM and 3.0 mM, respectively. Similar to GRSO 2− 4
and GRCl− , the GRHCO − also had potential for nitrate reduction 3 [36]. A green color was observed after settling during experiments in the system with externally adding Cl− , SO4 2− and HCO3 − . The introduction of 1.0 mM and 3.0 mM CH3 COO− achieved 88.25% and 93.4% nitrate removal, respectively. As a whole, the positive effects of the four anions above decreased in the order: CH3 COO− > SO4 2− > Cl− ≈ HCO3 − . Although the first ion was organic and the others were inorganic, they exhibited similar variation in pH and EC during the whole reaction period. The solution pH increased and EC decreased significantly within the first 40 h, and then kept stable (Fig. 3a, b, c, d). Both alkaline release from iron corrosion and acidity consumption from nitrate reduction by Fe0 contributed to pH increase. Nitrate reduction and subsequent ammonium adsorption, the adsorption and co-precipitation of other ions onto iron hydroxides all contributed to EC decrease. It implied that other co-existent contaminants (e.g., heavy metals) would be removed simultaneously by adsorption and precipitation. The nitrate removal rate was 62.4% and 60.9% (Fig. 3A, B) in the system externally adding 1.0 mM and 3.0 mM PO4 3− , respectively. The nitrogen recovery was the highest in all anions (data not shown). Phosphate has been reported to form co-precipitation and inner-sphere complexes on the surface of iron inhibited iron corrosion and electron transfer [37,38]. Although the addition of phosphate could increase ion strength and promote iron corrosion, the competitive adsorption was stronger. Oxalate achieved similar nitrate removal rate as Cl− in the same concentration (76.3% and 84.2%) (Fig. 3A, B). Among all anions, citrate achieved the best promotion to nitrate reduction by Fe0 . Almost complete removal of nitrate (>95%) was observed after 124 h in the presence of 1.0 mM and 3.0 mM citrate. The reactors containing citrate were very clear at the last period of the experiment, though they were turbid in the beginning of reaction as other systems. But the reactors containing
other anions were always turbid. Considering citrate was usually used as adsorbent or chelator in industry, it might have the same effect in this study. Citrate as a chelator could destabilize and finally dissolve the passive oxide layer on the surface of iron [37]. Although PO4 3− , oxalate and citrate behaved differently on nitrate reduction, they exhibited similar variation in pH and EC. The solution pH and EC varied repetitive and showed opposite to each other (Fig. 4). A possible explanation was that the added ion behaved as a buffer alone or together with some ions in soil solution, and thus they prevented the solution pH from fluctuating. The low pH led to dissolve the precipitates and/or desorb the adsorbed ions, which caused a rise of EC. Contrarily, the high pH induced precipitation/co-precipitation and then more adsorption, which decreased the EC. Consequently, the solution pH and EC changed simultaneously and exhibited a fluctuant situation, especially with higher concentration. In conclusion, all anions excluding PO4 3− revealed positive effects on nitrate reduction by Fe0 in the presence of loess soil. The promotion effect decreased in the order: citrate > acetate > SO4 2− > Cl− ≈ HCO3 − ≈ oxalate PO4 3− . Similar to cation, nitrogen recovery was only 61–78%. Some nitrite accumulation was observed in the beginning, but less than 1.0 mg N L−1 was detected after reaction in all systems with different anion present. 4. Fe0 -based PRB implication in the loess plateau area According to the above results, reductive removal of nitrate by Fe0 from an alkaline (pH > 8) soil was possible. The major anions (Cl− , SO4 2− , HCO3 − ) can accelerate nitrate reduction due to the promotion of iron corrosion and the possible occurrence of green rusts. The Fe0 corrosion products, Fe(II) and Fe(III), significantly enhanced nitrate reduction. Moreover, nitrate can be reduced to nitrogen gas by hydrogenotrophic denitrification process which consumed H2 deriving from iron corrosion [39,40]. More iron powder may be consumed in the regions where the concentration of PO4 3− is higher. Some organic ligands (e.g.,
C. Tang et al. / Journal of Hazardous Materials 231–232 (2012) 114–119
citrate, acetate and oxalate) would promote not only nitrate chemical reduction by Fe0 but also microorganism denitrification. For example, the denitrification rates decreased in the of order acetate > H2 > S > thiosulphate > ferrous iron in microcosm [41]. Certainly, there are lots of concomitant contaminants in the specific environment, of which the effects still need to investigate in detail. The iron in reactors after reaction was black and agglomeration. The passivation of iron surface resulting from high pH should account for the reaction expiration, which would decrease the longevity of Fe0 [42]. Thus, in an alkaline loess plateau area, more Fe0 consumption and relative short longevity of PRB may be the main concern. But it is still an alternative approach to remediate soil and groundwater nitrate contamination if no other better technologies to be developed in future. 5. Conclusions The results indicated that nitrate reduction by Fe0 in the alkaline loess plateau area was possible. High pH did not inhibit nitrate reduction reaction immediately. By contrast, the remove efficiency could be accelerated by some cations (Fe3+ , Cu2+ and Fe2+ ) and anions (citrate, acetate, oxalate, Cl− , SO4 2− and HCO3 − ). At the same concentration, the promotion effect of the cation decreased in the order of Fe3+ > Fe2+ > Cu2+ , but the anions decreased in this order of citrate > acetate > SO4 2− > Cl− ≈ HCO3 − ≈ oxalate PO4 3− . Only PO4 3− exhibited an inhibition for nitrate reduction by Fe0 . Ammonium was the major detectable final product from nitrate reduction. Nitrite, the intermediate product of nitrate reduction, was less than 1.0 mg N L−1 after 100 h reaction. But the nitrogen recovery in liquid and gas phase was only 56–78% in all systems. Dissolved iron was not detectable. The pH and EC varied depending on the character of added ion. This study implied that it is a potential technique to use Fe0 -based PRB for in situ remediation of nitrate contamination in the loess plateau area. Acknowledgements We acknowledge Mr. Su Chunming, National Risk Management Research Lab., U. S. Environmental Protection Agency, 919 Kerr Research Drive, Ada, Oklahoma, for editing the paper, and Dr. Brian D. Strahm, University of Washington, College of Forest Resources, Seattle, U.S., for correcting expression in the paper. We also thank the professor Yao Yaqin for SEM micrograph, College of Life Science, Northwest A & F University. We also acknowledge editor and anonymous reviewers for their suggestion and comments. References [1] Y. Zhang, Y.X. Chen, H.Y. Liu, Countermeasure and removal of nitrate in groundwater, Agro-Environ. Prot. 21 (2) (2002) 183–184 (in Chinese). [2] G.H. Jiang, W.K. Wang, Analysis of nitrate pollution of groundwater of Guanzhong basin and countermeasure, Water Resour. Prot. 68 (2) (2002) 6–8 (in Chinese). [3] J.O. Lundberg, E. Weitzberg, J.A. Cole, N. Benjamin, Nitrate, bacteria and human health, Nat. Rev. Microbiol. 2 (2004) 593–602. [4] S.M. Li, Survey and analysis on the main chemical and physical characters of water in Hongsipu irrigated area of Ningxia, J. Ningxia Agric. Coll. 24 (3) (2003) 37–39 (in Chinese). [5] X. Fan, X. Guan, J. Ma, H. Ai, Kinetics and corrosion products of aqueous nitrate reduction by iron powder without reaction conditions control, J. Environ. Sci. 21 (2009) 1028–1035. [6] S.M. Hosseini, B. Ataie-Ashtiani, M. Kholghi, Nitrate reduction by nano-Fe/Cu particles in packed column, Desalination 276 (2011) 214–221. [7] C.P. Huang, H.W. Wang, P.C. Chiu, Nitrate reduction by metallic iron, Water Res. 32 (8) (1998) 2257–2264. [8] Y.H. Huang, T.C. Zhang, Effects of low pH on nitrate reduction by iron powder, Water Res. 38 (11) (2004) 2631–2642. [9] D.W. Blowes, C.J. Ptacek, S.G. Benner, C.W.T. McRae, T.A. Bennett, R.W. Puls, Treatment of inorganic contaminants using permeable reactive barriers, J. Contam. Hydrol. 45 (2000) 123–137.
119
[10] J.F. Devlin, K.O. Allin, Major anion effects on the kinetics and reactivity of granular iron in glass-encased magnet batch reactor experiments, Environ. Sci. Technol. 39 (6) (2005) 1868–1874. [11] L. Xie, C. Shang, The effects of operational parameters and common anions on the reactivity of zero-valent iron in bromate reduction, Chemosphere 66 (2007) 1652–1659. [12] Y. Liu, T. Phenrat, G.V. Lowry, Effect of TCE concentration and dissolved groundwater solutes on nZVI-promoted TCE dechlorination and H2 evolution, Environ. Sci. Technol. 41 (2007) 7881–7887. [13] T. Liu, I.M.C. Lo, Influences of humic acid on Cr(VI) removal by zero-valent iron from groundwater with various constituents: implication for long-term PRB, Water Air Soil Pollut. 216 (2011) 473–483. [14] Y.H. Shih, M.Y. Chen, Y.F. Su, Pentachlorophenol reduction by Pd/Fe bimetallic nanoparticles: effects of copper, nickel, and ferric cations, Appl. Catal. B: Environ. 105 (2011) 24–29. [15] M. Hou, H. Wan, T. Liu, Y. Fan, X. Liu, X. Wang, The effect of different divalent cations on the reduction of hexavalent chromium by zerovalent iron, Appl. Catalysis B: Environ. 84 (2008) 170–175. [16] J. Xu, Z. Hao, C. Xie, X. Lv, Y. Yang, X. Xu, Promotion effect of Fe2+ and Fe3 O4 on nitrate reduction using zero-valent iron, Desalination 284 (2012) 9–13. [17] Y.H. Huang, T.C. Zhang, Enhancement of nitrate reduction in Fe0 -packed columns by selected cations, J. Environ. Eng. 131 (7) (2005) 603–611. [18] C.H. Liao, S.F. Kang, Y.W. Hsu, Zero-valent iron reduction of nitrate in the presence of ultraviolet light, organic matter and hydrogen peroxide, Water Res. 37 (2003) 4109–4118. [19] EPA, American Public Health Association, standard methods for the examination of water and wastewater, 19th, 1995. [20] S.F. Cheng, C.Y. Huang, J.Y. Liu, Study of different methods for enhancing the nitrate removal efficiency of a zero-valent metal process, Water Sci. Technol. 53 (2006) 81–87. [21] H. Huang, G.A. Sorial, Perchlorate remediation in aquatic systems by zerovalent iron, Environ. Eng. Sci. 24 (2007) 917–926. [22] C. Seunghee, M.L. Howand, Nitrate reduction by zero-valent iron under different pH regimes, Appl. Geochem. 19 (2004) 335–342. [23] C. Noubactep, A critical review on the process of contaminant removal in Fe0 –H2 O systems, Environ. Technol. 29 (2008) 909–920. [24] Y.J. Tsai, F.C. Chou, T.C. Cheng, Coupled acidification and ultrasound with iron enhances nitrate reduction, J. Hazard. Mater. 163 (2009) 743–747. [25] Q.J. Rasheed, K. Pandian, K. Muthukumar, Treatment of petroleum refinery wastewater by ultrasound-dispersed nanoscale zero-valent iron particles, Ultrason. Sonochem. 18 (2011) 1138–1142. [26] W. Kamolpornwijit, L.Y. Liang, G.R. Moline, T. Hart, O.R. West, Identification and quantification of mineral precipitation in Fe0 filings from a column study, Environ. Sci. Technol. 38 (2004) 5757–5765. [27] H.C.B. Hansen, G. Susanne, E. Marianne, Kinetics of nitrate reduction by green rust—effects of interlayer anion and Fe(II): Fe(III) ratio, Appl. Clay Sci. 18 (2001) 81–91. [28] Y.H. Huang, T.C. Zhang, J.S. Patrick, Effects of oxide coating and selected cations on nitrate reduction by iron metal, J. Environ. Qual. 32 (2003) 1306–1315. [29] Y. An, T. Li, Z. Jin, M. Dong, H. Xia, W. Xue, Effect of bimetallic and polymercoated Fe nanoparticles on biological denitrification, Bioresour. Technol. 101 (2010) 9825–9828. [30] Z.W. Hao, X.H. Xu, J. Jin, Simultaneous removal of nitrate and heavy metals by iron metal, J. Zhejiang Univ. Sci. B 5 (2005) 307–310. [31] Y.H. Hwang, D.G. Kim, Y.T. Ahn, C.M. Moon, H. Shin, Fate of nitrogen species in nitrate reduction by nanoscale zero valent iron and characterization of the reaction kinetics, Water Sci. Technol. 61 (2010) 705–712. [32] Y.H. Hwang, D.G. Kim, H.S. Shin, Mechanism study of nitrate reduction by nano zero valent iron, J. Hazard. Mater. 185 (2011) 1513–1521. [33] T. Fukuda, T. Mizuno, The evaluation of pitting corrosion from the spectrum slope of noise fluctuation on iron and 304 stainless steel electrodes, Corros. Sci. 38 (1996) 1085–1091. [34] H.Y. Hu, N. Goto, K. Fujie, Effect of pH on the reduction of nitrite in water by metallic iron, Water Res. 35 (2001) 2789–2793. [35] C.B.H. Hans, B.K. Christian, N.K. Hanne, Abiotic nitrate reduction to ammonium: key role of green rust, Environ. Sci. Technol. 30 (1996) 2053–2056. [36] C.M. Su, R.W. Puls, Significance of iron (II, III) hydroxycarbonate green rust in arsenic remediation using zerovalent him in laboratory column tests, Environ, Sci. Technol. 38 (2004) 5224–5231. [37] H. Harms, H.P. Volkland, G. Repphun, A. Hiltpolt, O. Wanner, A.J.B. Zehnder, Action of chelators on solid iron in phosphate-containing aqueous solution, Corros. Sci. 45 (2003) 1717–1732. [38] C.M. Su, R.W. Puls, Nitrate reduction by zerovalent iron: Effects of formate, oxalate, citrate, chloride, sulfate, borate and phosphate, Environ. Sci. Technol. 38 (2004) 2715–2720. [39] Y. An, T. Li, Z. Jin, M. Dong, Q. Li, Nitrate degradation and kinetic analysis of the denitrification system composed of iron nanoparticles and hydrogenotrophic bacteria, Desalination 252 (2010) 71–74. [40] S. Biswas, P. Bose, Zero-valent iron-assisted autotrophic denitrification, J. Environ. Eng. 131 (2005) 1212–1220. [41] J.F. Devlin, R. Eedy, B.J. Butler, The effects of electron donor and granular iron on nitrate transformation rates in sediments from a municipal water supply aquifer, J. Contam. Hydrol. 46 (2000) 81–97. [42] T. Kohn, K.J.T. Livi, A.L. Roberts, P. Vikesland, Longevity of granular iron in groundwater treatment processes: corrosion product development, Environ. Sci. Technol. 39 (2005) 2867–2879.